At Mnajdra, on Malta’s southern coast, the first light of sunrise reaches deep into a stone temple at the spring and autumn equinoxes. It passes between two great limestone uprights and follows a line laid out by builders more than five thousand years ago. At the summer and winter solstices, when the sun reaches its most northerly and southerly rising positions, the light falls instead on particular stones beside the inner passage. We can still see these effects today, although the building has lost its roof and much of its original surroundings.
Mnajdra belongs to a group of megalithic complexes built by farming communities on Malta and neighbouring Gozo in the fourth and third millennia BCE. Their curved façades lead into paved passages with chambers opening off a central route. These rounded side rooms are known to archaeologists as apses. Doorways, passages and chambers were arranged to shape the experience of entering the buildings and moving through them. Some openings also give a carefully framed view of the horizon. Whether the builders meant to capture particular appearances of the sun or stars is a question that the surviving stonework allows us to investigate.
The answer differs from temple to temple. At Mnajdra South, seasonal patterns of sunlight can be observed directly. Elsewhere, a doorway facing a particular part of the sky does not necessarily make a solar calendar. Research published in 2025 identified significant patterns among some of the earlier temple orientations, with bright southern stars emerging as possible targets. Many other orientations are also compatible with the lie of the land, prevailing winds, winter sunshine or chance. The evidence points to people who knew their landscape and watched the sky, but they left no written account of how they chose a building’s direction or what a sunrise meant to them.
The people who built Malta’s temples
The Maltese islands lie in the Mediterranean south of Sicily, with open water between them and the larger island. Their prehistoric communities farmed, kept animals, worked stone and maintained connections by sea. Temple construction unfolded over hundreds of years. Buildings were extended, rebuilt and sometimes used differently, rather than all being erected as parts of a single master plan. Archaeologists trace those changes through pottery and the sequence of the structures themselves. The Ġgantija phase, around the middle of the fourth millennium BCE, came before the later Tarxien phase, when some monuments grew and carved decoration became more elaborate. The chronologies differ in detail, and one building may preserve several phases of work.
Local limestone supplied the great upright slabs, or orthostats, that formed many of the walls and entrances. Builders laid lintels over doorways, paved central passages and made thick walls from large stones backed by smaller material and earth. Overlapping courses of masonry probably formed parts of the original roofs. From a façade that curves inward, a visitor often entered a passage flanked by rounded chambers. There is a recognizable architectural tradition here, but the monuments are far from identical. The UNESCO description of the Maltese megalithic temples stresses their individual differences. A sunrise measured at Mnajdra therefore cannot tell us, by itself, where the builders of every other Maltese temple meant their doorways to face.
Even the word temple carries an assumption. Stone altars, carved designs, figurines and traces of collective activity suggest ceremonies, feasting and shared identities, but no surviving Neolithic Maltese writing names a deity, priesthood or festival. For that reason, some archaeologists prefer to think in terms of communal or ceremonial houses rather than buildings equivalent to later Greek or Roman sanctuaries. These same communities also made collective burial places underground, most notably the Ħal Saflieni Hypogeum. Their monumental spaces evidently formed part of how the living related to one another and to their dead. What took place within them, and the beliefs behind those actions, are much harder to establish.
Why the horizon matters
Anyone who watches the sunrise from the same place over a year can see it move along the horizon. From Malta, its summer-solstice position is furthest to the northeast and its winter-solstice position furthest to the southeast. Around the spring and autumn equinoxes, when the sun crosses the celestial equator, it rises close to east if the horizon is level. Between these points its rising position changes from day to day. None of this requires written records or later mathematical astronomy to notice. Repeated observations from a familiar field, shore or temple forecourt would have shown a prehistoric community the same annual movement.
Establishing that a temple was aligned with one of those appearances is more demanding. Researchers measure its azimuth, a direction expressed in degrees clockwise from true north, and the altitude of the horizon in that direction. A hill delays the first visible appearance of the sun and changes its apparent position; atmospheric refraction also affects what can be seen. Stars present another difficulty. Through precession, the slow change in the orientation of Earth’s rotational axis, their rising positions shift over millennia. A star seen near one horizon point today may have appeared elsewhere when Malta’s temples were built. The relevant comparison must use a reconstructed ancient sky and the actual local horizon.
The architecture raises questions of its own. Researchers must decide whether to measure a central doorway, a side opening or a slanting path of light between two door jambs. Damage and rebuilding may have changed the outline of an entrance, while a roof that once blocked sunlight may no longer survive. A photograph of the present ruin cannot settle how an ancient observer experienced it. Sound interpretations therefore need measured stonework, a reconstructed horizon, a credible building date and a reason to think the proposed event was visible in its original setting. A close astronomical match makes a hypothesis worth testing. It cannot, on its own, establish the builder’s intention.
Mnajdra South and the equinox sunrise
Mnajdra stands on a terrace below the neighbouring complex of Ħaġar Qim. Three distinct buildings, joined around a forecourt, make up the site. The small eastern building, sometimes called the upper temple, came first; the South Temple was added later, followed by the central building on its raised platform. Their walls and platforms preserve evidence of this sequence. It is the South Temple, with its inward-curving façade and monumental entrance leading toward deeper chambers, that produces the best-known effects at sunrise. The separate building phases matter when interpreting observations made from different parts of the complex.
The South Temple does not point exactly east in a simple compass measurement. Frank Ventura and George Agius recorded its central axis at about 92°43′ clockwise from true north, slightly south of due east. The difference makes sense once the skyline is taken into account. Its eastern horizon stands several degrees above the observer, so equinox sunrise is first visible at a different bearing from sunrise over a flat sea. In their 2017 field study of the Mnajdra skyline, the researchers measured an altitude of roughly four degrees along the axis and calculated a solar declination close to zero. That combination explains how the temple can frame equinox sunrise even though its entrance is not directed at an azimuth of precisely 90 degrees.
The illumination is still visible to people attending special sunrise openings at Mnajdra. At the equinoxes, sunlight enters through the main doorway and travels along the South Temple’s central line. The beam moves as the sun rises, and its pattern changes over the days surrounding the equinox. Missing roof elements, possible changes to doorway height and unknown furnishings limit how exactly the ancient lighting can be reconstructed. The present evidence supports a recognizable seasonal effect. It does not show that the building indicated one uniquely defined calendar date with mechanical precision.
There is another distinction between the equinox as defined by astronomers and the date a Neolithic community might have considered important. Today the equinox is the moment when the sun crosses the celestial equator. A person observing from Mnajdra could instead have watched for sunrise at a notch on the skyline, judged when the interior light looked central, or counted an approximate interval between the solstices. Nothing on the surviving walls identifies which practice, if any, the builders followed. The entrance geometry and skyline fit the astronomical equinox closely, but they do not supply the missing rules of a prehistoric calendar.
What happens at the summer and winter solstices?
At the longest and shortest days, the morning light comes into Mnajdra South obliquely rather than travelling down the main axis. Around the summer solstice it catches the edge of a large pitted slab to the left of the central passage. At the winter solstice it reaches a decorated megalith on the right. The summer and winter beams strike separate interior stones, rather than opposite edges of one slab. Heritage Malta’s description of the summer event identifies the summer stone, while its winter account describes the winter effect. Because the accounts specify where the light falls, these observations can be checked against the building rather than accepted as a vague claim that the whole temple faces the sun.
A solstice is not easily pinpointed by watching the sunrise on one morning. Near its furthest northern or southern position, the sun changes its rising azimuth only slowly, so successive dawns can look very similar. The doorway’s width, the irregular edge of the stone, the apparent size of the solar disc and conditions in the atmosphere all affect the illuminated patch. Observers could nevertheless notice the repeated approach to, and retreat from, a seasonal extreme. What survives at Mnajdra is consistent with that kind of observation, but it cannot demonstrate that the builders identified the summer or winter solstice to the accuracy of a particular modern calendar day.
Heritage Malta, which manages the site, makes much the same distinction. In its account of the equinox and solstice events, the agency describes the repeating patterns as systematic and probably deliberate, but stops short of claiming certainty about the builders’ purpose. The light can be observed, although the dates, rites and decisions associated with it remain unknown. Dismissing the light as accidental would ignore the pattern. Treating the pattern as conclusive proof of an organized calendar of seasonal festivals would go beyond what the ruins and observations establish.
The possible horizon markers beyond the temple
Ventura and Agius also investigated the rocky skyline east of the temple to ask whether its builders had set out their axis using markers for the sun’s extreme rising points. They surveyed the horizon with a theodolite, an instrument for measuring horizontal and vertical angles, and tested possible marker positions against calculated solar declinations. Their 2017 fieldwork had practical obstacles. Some land was privately owned, and farming and bird trapping had disturbed sections of the limestone surface in more recent times. A hole that now lines up with a prehistoric sunrise might have been made centuries or millennia after the temple itself.
The team recorded a roughly circular hole about 35 centimetres wide and 36 to 37 centimetres deep. It could once have held a post, and their calculations put a post at that point very close to the winter-solstice sunrise direction around 3000 BCE. They could not establish when the hole had been made. They found a second hole in the region where a summer marker was expected, but it lay about three degrees away from the predicted direction. According to their estimate, the correct position would have been another 25 to 30 metres north. No matching feature was demonstrated there, leaving the proposed summer marker unsupported.
The surveyed holes thus offer a possible winter-solstice foresight, with no securely identified summer-solstice counterpart. Ventura and Agius presented this as a test of their hypothesis, not proof that the stones had been used to lay out Mnajdra South. Neither a timber post nor a securely dated Neolithic context survives, and no inscription explains a surveying procedure. Their recorded mismatch is as relevant as the apparent alignment. It shows how field archaeology can narrow the possibilities while leaving a plausible interpretation unresolved.
Ħaġar Qim and the summer-solstice aperture
Uphill from Mnajdra, the linked chambers and curved walls of Ħaġar Qim produce a different effect in the summer sunlight. At sunrise near the June solstice, light passes through a small elliptical aperture into a room on the northeastern side. On a stone surface inside, the illuminated shape begins as a crescent, expands into a disc and then shifts away as the sun rises. Heritage Malta documents the aperture and changing light shape in its account of the summer-solstice opening. Here the significant feature is a particular hole and the moving patch of light it admits, rather than the alignment of a main central passage.
An aperture can restrict sunlight to a small area even when the building as a whole does not face the sunrise that enters it. To understand what the effect originally meant, researchers must consider whether the surrounding walls and opening had their present arrangement when that part of Ħaġar Qim was made. The observation itself is specific; the date of the architectural arrangement and any deliberate significance assigned to it require separate arguments. It would be misleading to combine this aperture effect with the entrance bearings of dozens of other temples and treat them as identical evidence of solar design.
Do Ġgantija, Tarxien and other temples face the same sky?
Other major complexes share the Maltese tradition without reproducing Mnajdra’s particular display. Ġgantija on Gozo has two imposing temples with massive walls and rounded apsidal rooms. Tarxien on Malta preserves elaborate carving and a more complex history of additions and alterations. Both contribute valuable measurements to comparisons of temple orientations, but neither has an established equinox light sequence exactly equivalent to the one through Mnajdra South’s main doorway. At Tarxien especially, structures belong to different building phases. Combining their axes into one supposedly permanent Maltese preference risks overlooking how the tradition changed over time.
In earlier surveys, Agius and Ventura found that entrances tended to face sectors of the southeast and southwest horizon, rather than being distributed uniformly around the compass. The preferred sectors also differed between examples assigned to the Ġgantija and Tarxien phases. An entrance pointing southeast could have more than one explanation: shelter from wind, an accessible slope, useful winter daylight or the rising place of a star. Even a genuine pattern of directions therefore does not identify its cause. Any proposed astronomical explanation has to compete with what the landscape and the practical needs of builders already make plausible.
Robert Peter Barratt examined 23 Maltese temple structures in a study published in 2022. He used three-dimensional models and raycasting, which follows possible sightlines through a digital reconstruction, to test their orientations. His published abstract and institutional research record report a cautious preference for possible alignments involving Crux, the Southern Cross, over the solar explanations often assumed for the wider temple group. This result concerns the particular buildings and directions included in that analysis. It cannot undo the observed sunlight at Mnajdra, and it does not show that every doorway was constructed to observe the same constellation.
The 2025 statistical study and the southern stars
A larger comparison appeared in April 2025, when Fabio Silva and Tore Lomsdalen published their study of 32 temple structures. Their open-access research in Archaeological and Anthropological Sciences considered astronomical suggestions alongside other ways the buildings might have been oriented. The authors tested how the axes compared with random directions, slopes in the surrounding land, protection from wind and exposure to winter sunlight. They reconstructed the skyline seen from each structure and allowed for uncertainties in the surviving masonry and available landscape information. By testing these alternatives, the study made it possible to ask which apparent sky alignments remained unusual after more ordinary influences had been considered.
Many of the recorded directions were compatible with those non-astronomical explanations. Among the earlier buildings, however, some statistically significant patterns remained. When Silva and Lomsdalen calculated where celestial bodies would have appeared in the Neolithic sky, the possible matches included Hadar in Centaurus, Gacrux in the Southern Cross and Avior in Carina, a star associated with the False Cross. All three belong to the bright southern sky. Their ancient rising positions cannot be read directly from a present-day sky chart, because precession has altered the relationship between the stars and the horizon over the intervening millennia.
The authors did not equate a statistically unusual result with proof that a builder intended to sight a star. Some entrance lines are uncertain because stones are damaged, and some construction dates cannot be fixed closely. When the sample was separated into building phases, the significant results applied to the earlier Ġgantija-phase group, not to the later temples as a whole. Silva and Lomsdalen therefore did not propose a common stellar programme for every monument. Their results strengthen the case for possible stellar targets among some early temples, but offer no support for a single solar calendar encompassing the whole group.
One reason southern stars might have mattered is seafaring. Silva and Lomsdalen noted that travellers sailing north from Malta could sometimes use the high ground of Sicily as a visual guide in clear weather. Returning south, they faced the harder task of locating low-lying Malta after it had disappeared below the horizon. A succession of rising stars might have helped them maintain a direction. The researchers even suggested that some temples could have been places where younger voyagers learned these patterns. They identified this teaching function as speculation. No Neolithic navigator’s account, training manual or excavated model boat establishes it. The proposed connection between star alignments and the practical demands of island seafaring remains open to archaeological testing.
The drilled stones at Mnajdra and a possible record of time
Mnajdra’s older eastern temple contains another feature that may be relevant to recording celestial time. Pillars near an inner apse carry rows of small drilled holes, and some researchers have suggested that their numbers represent intervals between particular stellar appearances. A star may disappear into the dawn glare for a period and then become visible again briefly before sunrise. Astronomers call that first pre-dawn reappearance a heliacal rising. Because it tends to recur at a particular season, it can be used to mark the passage of the year. The question is whether the makers of the holes were recording such intervals or doing something unrelated.
In their 2021 statistical study of the Mnajdra tally, George Agius, Lorraine Brown Read and Frank Ventura compared the surviving hole counts with calculated intervals between heliacal risings of bright stars around 3000 BCE. Their statistical tests suggested that the correspondence was unlikely under the random orderings they examined. That is evidence worth considering, but it does not amount to reading a prehistoric calendar. Some holes are difficult to count because of erosion, the reason for the proposed sequence is unclear, and the pillars carry no explanation of the markings. They might have counted days, but decoration, accounting or some other activity has not been ruled out.
Another possibility concerns the Pleiades, the prominent star cluster in Taurus. Its calculated rising direction in the ancient sky could have been close to Mnajdra South’s equinox axis. In much later societies, appearances of the Pleiades served as seasonal signs, although that comparison cannot establish the intentions of Neolithic Maltese builders. It does point to a problem with assigning one and only one target to a doorway. The same architectural line may admit equinox sunlight at one time and frame a star cluster at another. The observed solar match is real; deciding whether the builders also cared about the stars requires evidence that the alignment alone does not supply.
What astronomy can and cannot tell us about ritual life
A distinctive sunrise entering a ceremonial building could have been an occasion for people to gather. Farming communities had reasons to notice seasonal change, and the deliberate arrangement of entrances and internal spaces may have made repeated light effects meaningful. Yet the archaeological evidence does not show that every farmer could obtain an exact agricultural date from an illuminated wall, or that a priest used the beam to announce planting and harvest. Rainfall, soil, crops and available labour all affected agricultural work, as well as the length of the day. Without records of the ceremonies themselves, an attractive reconstruction of ritual practice must remain a possibility rather than a reported fact.
The Ħal Saflieni Hypogeum shows how much care these communities could give to experiences within a building. Its underground burial chambers, carved entrances and red-ochre designs were created over generations. One space, called the Oracle Room, is known for its striking acoustic resonance. Heritage Malta’s account of the Hypogeum describes both the resonant niche and the site’s funerary character. The Hypogeum cannot prove that the same people designed astronomical calendars above ground. It does show that prehistoric architecture could shape sound, darkness, access and movement as well as the play of light. The direction of a doorway was only one part of the experience its builders could arrange.
It is therefore difficult to treat every Maltese megalithic building as an observatory in the modern sense of a specialized place for measuring the heavens. The monuments remained in use and changed across long periods; their axes may reflect astronomy, local terrain, inherited building practice and community choices. A striking solar illumination at Mnajdra could have had a ritual function. It still tells us little about who entered, who stayed outside, whether anything was offered, or how a ceremony may have changed from one generation to the next. Those questions belong to the history of the communities, and they cannot be answered from a solar beam alone.
The individual findings need to be kept distinct. Mnajdra South still receives equinox sunlight along its axis and solstice light on two separate stones. Proposed postholes on the eastern horizon have not established how the builders set that axis out. Ħaġar Qim has its own summer-solstice aperture effect. Surveys of other temples show that neither sunlight nor any single celestial explanation accounts for all their directions, while the 2025 analysis makes bright southern stars plausible targets for certain earlier structures. The original ceremonies remain unknown, but the monuments preserve enough of their shape and relationship to the landscape to reconstruct some of the sky seen by communities in their fields, along the coast and within temple forecourts.
Discussion
Comments